1. Forum
  2. FidoNet
  3. FIDOSOFT.HUSKY

  • loglevels

    From Tommi Koivula@2:221/360 to All on Sat Feb 27 08:48:54 2021

    Hi All,

    I installed openwatcom 2.0 for OS/2. Compiled, and hpt seems to run fine. Thanks Andrew! :)

    I turned on logging "1" and now the log is full of "scanning area" lines starting with "1".

    Is this ok? According to the docs :

    1 - Program start, end;
    4 - scanning messages

    === Cut ===
    ---------- Sat 27 Feb 2021, hpt/os2-wc 1.9.0-cur 2021-02-17
    1 08:46:51 Start
    1 08:46:51 Start scanning...
    1 08:46:51 EchoTossLogFile not found -> Scanning all areas
    1 08:46:51 Scanning area: r20_kontroll
    1 08:46:51 Scanning area: r20_test
    1 08:46:51 Scanning area: r20_admin
    1 08:46:51 Scanning area: r20_echo2
    1 08:46:51 Scanning area: r20_beer
    1 08:46:51 Scanning area: r20_best_of_fidonet
    1 08:46:51 Scanning area: r20_chat
    1 08:46:51 Scanning area: r20_dev
    1 08:46:51 Scanning area: r20_info2
    1 08:46:51 Scanning area: r20_nc
    1 08:46:51 Scanning area: r20_operativsystem.linux
    === Cut ===

    'Tommi

    ---
    * Origin: rbb.fidonet.fi (2:221/360)
  • From Michael Dukelsky@2:5020/1042 to andrew clarke on Sat Feb 27 18:03:02 2021
    Hello andrew,

    Saturday February 27 2021, andrew clarke wrote to Tommi Koivula:

    $ cd hpt/src
    $ grep "Scanning area" *.c
    scanarea.c: w_log(LL_START, "Scanning area: %s",
    echo->areaName); toss.c: w_log(LL_START, "Scanning area: %s", config->badArea.areaName);

    At first glance this a bug, and "LL_START" should be "LL_SCANNING".

    I also think this is a bug. LL_START should be used at the beginning of a function and before every "return" from it.

    Michael

    ... node (at) f1042 (dot) ru
    --- GoldED+/LNX 1.1.5-b20180707
    * Origin: Moscow, Russia (2:5020/1042)
  • From Kai Richter@2:240/77 to andrew clarke on Sat Feb 27 16:19:12 2021
    Hello andrew!

    27 Feb 21, Michael Dukelsky wrote to andrew clarke:

    scanarea.c: w_log(LL_START, "Scanning area: %s",

    At first glance this a bug, and "LL_START" should be "LL_SCANNING".

    I also think this is a bug. LL_START should be used at the beginning
    of a function and before every "return" from it.

    Hm, from what the user documentation said i would expect level 1 to record the start of a module (hpt, htick, hptsqix, nltool...). More LogLevel is needed for troubleshooting only. Once the software is working well, i would like to see a confirmation that it was started to make sure previous scripts did their job.

    Are there other locations in the source where LL_SCANNING is used to control output?

    The husky loglevels system is other that software that defines a single loglevel which includes all lower LL. Husky does it more detailed like "turn level X on/off seperately". May that's why it's called LogLevel*s* and not LogLevel, i didn't noticed that fine difference in the keyword before.

    Regards

    Kai

    --- GoldED+/LNX 1.1.4.7
    * Origin: Monobox (2:240/77)
  • From Michael Dukelsky@2:5020/1042 to Kai Richter on Sun Feb 28 11:30:56 2021
    Hello Kai,

    Saturday February 27 2021, Kai Richter wrote to andrew clarke:

    scanarea.c: w_log(LL_START, "Scanning area: %s",

    At first glance this a bug, and "LL_START" should be
    "LL_SCANNING".

    I also think this is a bug. LL_START should be used at the
    beginning of a function and before every "return" from it.

    Hm, from what the user documentation said i would expect level 1 to
    record the start of a module (hpt, htick, hptsqix, nltool...).

    Yes, you are right, I was too hasty, "LL_FUNC" should be used for entering and leaving a function but still "LL_START" should not be used instead of "LL_SCANNING".

    Michael

    ... node (at) f1042 (dot) ru
    --- GoldED+/LNX 1.1.5-b20180707
    * Origin: Moscow, Russia (2:5020/1042)